Understanding the self-care behaviors by which elderly persons maintain independence is critical in planning long-term care services. Arthritis is among the most prevalent and disabling chronic conditions faced by elderly populations, and requires much personal involvement by those afflicted to optimize daily functioning. Given a rapidly growing elderly population, even small decreases in average rates of institutional care can substantially reduce national health care costs. The proposed study draws upon a new national sample database of 3,485 community-dwelling elderly persons, including 2,043 subjects with arthritis, to examine self-care practices related to arthritis and joint pain. This is the first nationally-representative sample establishing prevalence of self- care in the non-institutionalized elderly population, and linking self- care activity to national prevalence data on arthritis and other chronic conditions, functional status, social support, and health risk factors. The database includes (1) baseline in-person interviews (1990-1991); (2) follow-up telephone interviews (to be collected in 1993 in conjunction with a newly-funded NIA study); (3) health services use data from Medicare beneficiary files; and (4) National Death Index records. We will examine baseline relationships of arthritis (including arthritis diagnosis) and self-care with socio-demographic factors, health status and co-morbidity, functional status, social support, social-psychological factors, and medication use; and longitudinal relationships of these with follow-up outcomes: functional and health status, death, health service use, Medicare costs, and institutionalization. The sample frame used Health Care Financing Administration (HCFA) records for non-institutionalized Medicare beneficiaries who were 65 years and older. The stratified random national sample is geographically-clustered within 50 randomly selected sampling units in 38 urban and 12 rural areas across the United States. The design over-sampled oldest ages, assuring representation of those most likely to be frail and difficult-to-reach. Weights are applied to achieve a nationally-representative sample of well and frail elderly persons for age, gender and racial groups in rural and urban areas. Despite potential importance of self-care as part of a national strategy for health enhancement and health care cost-containment for the growing elderly population, virtually nothing is known about the types and extent of self-care activity by elderly persons for managing arthritis, its pain and functional limits. The goals of the original project for which data were collected did not focus specifically on arthritis. Similarly, the NIA-funded follow-up that is planned was not originally designed to ask arthritis-specific questions. However, the proposed project makes use of the existing data and the forthcoming follow-up data collection activity to address self-care practices related specifically to arthritis among the elderly group surveyed. Proposed study findings should help clinicians understand patients' perspectives of arthritis and to identify the relative importance of various self-care strategies, health care, and other factors that facilitate patients' effects to manage arthritis and maintain independence. Such information should better prepare physicians to intervene to help their patients understand and practice appropriate, safe arthritis self-care, and reinforce compliance with recommended medical care in daily activity.